Secondary Logo

Journal Logo


Oral Ketamine/Midazolam Is Superior to Intramuscular Meperidine, Promethazine, and Chlorpromazine for Pediatric Cardiac Catheterization

Auden, Steve M. MD*; Sobczyk, Walter L. MD; Solinger, Robert E. MD; Goldsmith, L. Jane PhD

Author Information
doi: 10.1213/00000539-200002000-00011


Forty years ago, Smith et al. (1) described sedation of children for cardiac catheterization (cath) by using “an ataractic mixture.” In their series of 670 patients, they reported that “only 17 appeared shocked,” and that 25% were “restless.” In addition, they reported three episodes of serious respiratory depression and one death, caused in part by this medication. Despite these shortcomings, their “cardiac cocktail,” a mixture of meperidine, promethazine, and chlorpromazine (DPT) filled a need at the time. IM DPT became the standard sedative for pediatric cardiac cath and for a variety of other procedures in children.

IM DPT continues to be widely used and, at least lately, widely reviled. Major shortcomings of DPT include its painful route of administration, slow onset, prolonged effect, lack of reliable amnesia, and frequent occurrence of restlessness. Respiratory depression is common and respiratory arrest can occur (2), with two prospective series putting the incidence of serious to life-threatening complications at 4% (2,3). Respiratory depression often reflects excess sedation; however, deep sedation is not reliably achieved with DPT. Terndrup et al. (4) reported a 29% failure rate for emergency department procedures. Prolonged duration of sedation from DPT was also reported in the same study, in which 19 ± 15 h passed before return to normal behavior (4). These and other problems have led to calls for “rational and safe alternatives” (5), and the American Academy of Pediatrics Committee on Drugs has issued a critical “reappraisal of lytic cocktail” (6).

For some time, we had been using an oral (PO) combination of ketamine/midazolam as premedication for cardiac surgery in children. We had found the combination of ketamine/midazolam to be free of the negative side effects of ketamine alone (7–9) and to provide rapid onset of deep sedation with minimal, if any, hemodynamic or respiratory compromise (10–11). Both components confer amnesia. Accordingly, we evaluated PO ketamine/midazolam as an alternative to IM DPT as the primary sedative for pediatric cardiac cath.


After institutional review board approval and written, informed consent, 51 children, ages 9 mo to 10 yr, scheduled for elective cardiac cath were enrolled and completed the study. During the precatheterization clinic visit, consent was obtained; height, weight, and oxygen saturation were recorded; and patients were assigned to diagnostic groups and designated as cyanotic or acyanotic by the attending cardiologist. Because of the varied clinical conditions and diagnoses of children coming for cardiac cath, randomization was done within each of 12 diagnostic categories. Patients with atrial septal defects were thus randomized separately from those with single ventricle, etc. Within each diagnostic group, children were randomized to receive either IM DPT or PO ketamine/midazolam.

On the day of the scheduled procedure, children came to the preoperative area. Before any medication was given, vital sign measurements were made including heart rate (HR) by electrocardiogram, mean blood pressure (BPM) by oscillotonometry, and SpO2 by pulse oximetry. These measurements were made by using a portable Escort® monitor (Medical Data Electronics, Arleta, CA) which remained with the patient and was used throughout the study. Respiratory rate (RR) was counted for a minimum of 15 s. An observational sedation scale was assessed (and later repeated as noted) by using a scale of awake, drowsy, or asleep. Any patient assessed as drowsy or asleep after drug administration was considered sedated. When possible, an echocardiogram was obtained (33 of 51 patients, with two studies later excluded for inadequate quality). Ventricular shortening fraction (SF) was determined by measurements of recorded echoes at a later date. SF was measured from a standard parasternal long-axis view.

After these assessments, IM injection was given to all patients, with time of the IM injection defined for study purposes as time zero. Acyanotic patients in the IM group received IM DPT at a dose of 2 mg/kg of meperidine and 1 mg/kg each of promethazine and chlorpromazine. Patients in the PO group received IM saline at time zero. The IM medication dose (and the volume of the IM placebo) was one-half for cyanotic patients, which is our common practice and reflects the usual practice elsewhere (1,12).

All patients received PO fluid at time = 15 min (i.e., 15 min after IM injection, to blind for the slower onset of IM DPT). Whereas the IM dosage was adjusted based on cyanosis versus acyanosis, the PO dosage was adjusted to give younger patients a larger dose (13,14). Children in the PO group who were ≤3 yr old received ketamine 10 mg/kg and midazolam 1 mg/kg. Children ≥4 yr old received ketamine 6 mg/kg and midazolam 0.6 mg/kg (15–19). Children in the IM group received a placebo of the flavored vehicle only, volume adjusted by age. All drugs and placebos were given in double-blinded fashion. All randomization and drug/placebo dispensing was performed in the pharmacy and all other personnel were completely blinded. Patient acceptance of the IM and PO medications was rated on a three-tier observational scale of marked response versus moderate response versus minimal or no response (See Table 1).

All patients were closely observed and monitored. Observable onset of sedation (i.e., the time at which the patient was first noted to be drowsy) and onset of sleep were noted for all patients. At minutes 25, 35, and 45, vital signs (as previously described) were repeated. After the 45-min vital signs, 1% carbonated lidocaine (20) was injected to allow placement of an IV cannula. Patient response to local anesthetic was assessed, again by using the observational scale of marked response versus moderate response versus minimal or no response (See Table 1).

After securing IV access, patients were transported to the cardiac cath lab where the echocardiogram was repeated. Children were then separated from their parents and moved to the cardiac cath table where they were positioned, prepared, and draped. Local anesthetic was infiltrated over the femoral vessels to allow cannulation. At separation, positioning, and cannulation, patient response was once again assessed by using the three-tier observational scale described in Table 1. At positioning and cannulation and for the remainder of the procedure, the attending anesthesiologist administered additional sedative medication based on patient responses. In all cases, the additional medication was propofol 0.5 mg/kg given IV via pump over 30 s (21). The patient was observed for cardiorespiratory effects of propofol, and any 10% change in BPM or 6-point change in SpO2 was recorded. Doses of propofol were tracked, and when more than 10 boluses were used, it was at the anesthesiologist’s discretion to begin a propofol infusion at 100 mcg/kg/min. The infusion could be adjusted by 25 mcg/kg/min increments as needed.

Table 1
Table 1:
Observational Scale for Patient Responses

Time spent in the recovery area was also noted. Our routine recovery discharge criteria were used to determine the stay.

After a minimum of 4 h on the floor, a final interview with patients and parents was conducted. Parents completed a visual analog scale as a measure of satisfaction with sedation. A 10-cm unmarked line labeled “failed sedation” at the left end [0] and “excellent sedation” at the right end [100] was presented. Parents marked their assessment, and these marks were later given a numeric value by ruler measurement. Children ≥4 yr old were questioned as to recall of IV placement or events in the cardiac cath lab. Any positive response was judged as failure to provide amnesia.

Statistical analysis was performed by using the Mann-Whitney U-test for non-Gaussian data and Student’s t-test or repeated measures analysis of variance for Gaussian data. Binomial data were analyzed by using the Pearson χ2 or Fisher’s exact test, as appropriate. Statistical significance was defined as P ≤ 0.05.


There were no intergroup differences for age, weight, time to cannulation, or duration of procedure (Table 2). As expected, children preferred oral medication, 49 of 51 patients showing better acceptance of PO fluid versus IM injection (P < 0.0005).

Table 2
Table 2:
Observations, Measurements, and Between Group Comparisons

In both groups, a high percentage of children became drowsy, with no significant difference between groups (22 of 23 in the PO group, 28 of 28 in the IM group, P = 0.451, Fisher’s exact test). The exception was one child who had expectorated a portion of the PO medication. Likewise, most children fell asleep, with 20 of 23 children in the PO group and 23 of 28 in the IM group asleep in the first 45 min of observation (again, no significant difference between groups, P = 0.715, Fisher’s exact test). Onset of observable sedation (drowsiness) and sleep from time zero were similar between groups. Observable sedation occurred at 28 ± 5 min (PO) and at 25 ± 9 min (IM), and onset of sleep was at 35 ± 15 min (PO) and 33 ± 12 min (IM). When the 15 min delay before administration of PO ketamine/midazolam is taken into account, onsets of both sedation (13 min PO group versus 25 min IM group, P < 0.0005) and sleep (20 min PO group versus 33 min IM group, P = 0.001, Mann-Whitney U-test) were significantly earlier in the PO group. Although the overall incidences of sedation and of sleep were similar between groups, patient responses to stimulation varied markedly. For example, on arrival to the cardiac cath lab and transfer from stretcher, 17 of 23 patients in the PO group remained asleep, five were drowsy, and only one was awake. Only 8 of 28 patients in the IM group remained asleep, whereas 13 were drowsy and 7 were awake (P = 0.004, Pearson χ2). Patients who received PO ketamine/midazolam tolerated IV placement, separation from parents, and femoral cannulation more readily than their counterparts in the IM group (Figure 1). Patients who had received IM DPT were more restless and tended to awaken and often became agitated with movement from stretcher, positioning, subcutaneous injection of local anesthetic, etc. Thirteen children in the IM group were noted to be agitated at one or more points as opposed to only two children in the PO group (P = 0.0048, Fisher’s exact test). There were no differences for onsets of sedation or sleep for cyanotic versus acyanotic children (P = 0.62 and 0.18, respectively, Mann-Whitney U-test).

Figure 1
Figure 1:
A, Response to local anesthetic injection for placement of a peripheral IV line. Asterisk (*) denotes zero patients in the given response category. By Pearson χ2 testP < 0.0005 between im and oral administration (PO) groups. All y axes refer to number of patients with given response. B, Response to separation from parents. Asterisk (*) denotes zero patients in the given response category. By Pearson χ2 test P < 0.0005 between im and oral administration (PO) groups. C, Response to local anesthetic and placement of femoral cannulae. By Pearson χ2 test P < 0.0005 between IM and oral administration (PO) groups.

There were no significant changes in HR over time or between groups (Figure 2). After administration of medication, SpO2 had a small decrease, the mean SpO2 difference from baseline to arrival in the cardiac cath lab being −1.5% (95% confidence intervals −0.6 to −2.4, P = 0.002 by Student’s t-test). BPM was also lower over time (P = 0.015, repeated measures ANOVA). Neither SpO2 nor BPM data showed any difference between groups. The initial RR in the two groups were essentially identical (P = 0.981). The RR for the two groups differed over time (P < 0.005, repeated measures ANOVA). Graphing the RR reveals that this difference is primarily caused by a decrease in the IM group (See Figure 2). Other cardiorespiratory changes in the IM group included two patients with >10% drop in BPM and one patient with a drop in SpO2 requiring supplemental oxygen. One patient required jaw thrust and another required oxygen by nasal cannula for decreased SpO2, both in the PO group.

Figure 2
Figure 2:
Changes in cardiorespiratory variables are shown over time from before medication baseline (BASE) and at 25, 35, and 45 (T25, T35, T45) minutes after IM injection. These times coincide with 10, 20, and 30 min after PO medication. The final measurement is on arrival (ARR) in the cardiac catheterization lab. For all variables, the dashed line represents the PO group and the solid line the IM group. Error bars reflect SD and are shown as an open bar for the PO group and a “T-bar” for the IM group. Variables include heart rate (HR) in bpm, percent oxygen saturation (SpO2), mean blood pressure (BPM) in mm Hg, and respiratory rate (RR) in breaths/min. + =P < 0.05 versus baseline by repeated measures of analysis of variance, and * =P < 0.05 between groups.

Echocardiography revealed a wide variation in SF after medication, however, there was no clear difference between treatment groups. The overall effect was a trend toward decrease in SF, which did not achieve statistical significance. Mean change in SF was −1.5% from before to after medication (95% confidence intervals + 0.9% to −3.9%, Student’s t-test for difference, P = 0.21).

Supplemental analgesia and sedation with propofol were required by a higher percentage of patients in the IM group (See Table 2). This was true on transfer to the cardiac cath table (on arrival), on femoral cannulation, and throughout the procedure. The median number of doses of propofol in the IM group was one dose on arrival, two with cannulation, and six total. For the PO group, the median number of doses of propofol was zero both on arrival and cannulation, and one total (P = 0.046 arrival, 0.01 cannulation, 0.002 total; Mann-Whitney U-test). With bolus use of propofol, two patients (both in the IM group) had >10% decrease in BPM and two others (both in the PO group) had SpO2 decrease transiently, responding readily to jaw thrust.

Recovery data reflects 48 patients. Of the other three, one patient in the IM group went to the operating room, one patient in the PO group went to intensive care, and one patient received fentanyl in recovery for hypercyanotic episodes. Recovery room discharge was delayed in the IM group (See Table 2). Despite their longer stay in recovery, patients in the IM group were more sedated at time of discharge than those in the PO group (P = 0.001, Pearson χ2).

Parental satisfaction was greater with the PO medication (See Table 2). In children ≥4 yr old, only one child in the PO group had recall (of IV placement). More than one-half of the patients in the IM group had recall of IV placement and/or the cardiac cath lab (See Table 2).


This study reveals many advantages of PO ketamine/midazolam over IM DPT: better accepted route of administration, faster onset, better toleration of minor procedures, less agitation, less need for supplemental sedation, better amnesia, and improved parental satisfaction. Recovery is more rapid and predictable. Hemodynamic responses to both medication regimens were comparable: HR was stable, SpO2 change was minimal, and the slight decrease in BPM was comparable to that with normal sleep (22). RR decreased only in the IM group. A decrease in RR is expected with narcotic medications such as meperidine, and unchanged RR has been previously described in children receiving ketamine (23) or nasal ketamine/midazolam (11). However, RR may not reflect more subtle alterations in ventilatory responses (23) or changes in ventilation caused by sedation. For example, two patients needed some type of respiratory support after PO ketamine/midazolam. These children maintained RR, however, they were sedated enough that decreased glossopharyngeal tone impaired oxygenation. Also, two additional patients in the PO group needed temporary airway support (jaw thrust) after propofol bolus. Airway support, a problem with DPT, may also be needed after PO ketamine/midazolam or after propofol, and neither ketamine/midazolam nor propofol is a panacea for the problems of pediatric sedation.

There are two interrelated problems with applying the results of this study to routine cardiac cath lab procedure: staffing and safety. Smith et al. (1) popularized their “ataractic mixture” for pediatric cardiologists who then monitored the patient, performed the procedure, and gave additional sedative medication as needed. This division of attention is no longer acceptable under modern sedation guidelines (24,25). In our study, each child had a pediatric anesthesiologist in attendance. Additional sedation, i.e., propofol, was given to maintain a level of sedation acceptable to both the cardiologist and the anesthesiologist. The quality of sedation, the supplemental use of propofol, and the ease of dealing with complications are all no doubt affected by the presence of an anesthesiologist. We found that supplemental use of propofol was less frequent in the PO group. This was true both as a percentage of patients requiring medication and as median number of doses required, and underscores the greater efficacy of PO ketamine/midazolam.

Secondly, our study population was too small to estimate safety or complication rate of either ketamine/midazolam or of supplemental propofol, however, minor complications did occur. By power analysis, if we accept a DPT rate of serious complications as 4%, and seek to differentiate this from a 10% rate of problems with our PO combination, we would need to study 300 patients in each group. If the PO complication rate is 1%, 450 patients would be required in each group to define this difference. This issue of safety is inextricably entwined with the staffing issue previously mentioned, and with the issue of who should administer sedative medications. In many hospitals, both ketamine and propofol are classed as anesthetic drugs, with their use restricted to anesthesiologists. Acute airway obstruction in an 11-month-old infant may be a life threatening event for a cardiologist, however, it would be a routine and easily managed event for the anesthesiologist.

When this study was designed in 1994, an informal survey revealed that 60% of pediatric cardiac cath labs routinely used DPT. In 1995 the AAP Committee on Drugs (6) published its critique of DPT, stating in the conclusions that whereas “… DPT … remains a widely used sedative and analgesic…. Neither the combination itself nor its dosage is based on sound pharmacologic data. There is a high rate of therapeutic failure as well as a high rate of serious adverse reaction, including … death, associated with its use. Because of this critique and the reports of other complications (1–6), alternatives should be sought. Clearly, within the framework of this study, PO ketamine/midazolam was superior. The framework of this study, however, included the presence of a pediatric anesthesiologist in the cardiac cath lab. In our own hospital, we have seen more anesthesia involvement with cardiac caths. This is partly related to the increased number of interventional procedures, however, it also reflects recognition of superior patient sedation seen in this study. Superior patient safety may also be a benefit, albeit an unproved one.

A properly blinded study design was a major concern. First, IM DPT may be more painful than IM saline, and can leave a “knot” at the injection site. The cardiac cath lab nursing team noted one 2 × 2-cm knot; however, these differences were otherwise not apparent. Similarly, we were concerned that PO ketamine/midazolam would be poorly tolerated compared with the flavored vehicle only; however, this was also not apparent. Only 1 of the 51 patients expectorated a portion of the PO fluid. Finally, problems with oral secretions, dysphoric reactions, and cardiosympathetic stimulation can occur when ketamine is given alone, including in children (7,26,27). None of these problems occurred with the combination of ketamine/midazolam. The combination of these two drugs also yielded sleep rather than the eyes-open, dissociated state that can occur with ketamine alone. Interestingly, two patients noted under “comments” to be “dissociated but restless” were both in the IM group. The staggered time frame of dosing, along with these findings, allowed successful blinding.

In summary, we found that PO ketamine/midazolam is clearly superior to IM DPT for pediatric cardiac cath. This regimen must be delivered safely, and safety may require that anesthesiologists play a role in the pediatric cardiac cath lab.

The authors wish to acknowledge the invaluable expertise of Victor Whalen, RPh, the assistance of David E. Miles, BS, and the administrative grace of Ms. Glynda Brooks. We would also like to thank Patty Welch, CCRN, who pushed us to begin and to pursue this investigation.


1. Smith C, Rowe RD, Vlad P. Sedation of children for cardiac catheterization with an ataractic mixture. Can Anaesth Soc J 1958; 5:35–40.
2. Nahata MC, Clotz MA, Krogg EA. Adverse effects of meperidine, promethazine, and chlorpromazine for sedation in pediatric patients. Clin Pediatr 1985; 24:558–60.
3. Mitchell AA, Louik C, Lacouture P, et al. Risks to children from computed tomographic scan premedication. JAMA 1982; 247:2385–8.
4. Terndrup TE, Dire DJ, Madden CM, et al. A prospective analysis of intramuscular meperidine, promethazine, and chlorpromazine in pediatric emergency department patients. Ann of Emerg Med 1991; 20:31–5.
5. Snodgrass WR, Dodge WF. Lytic/“DPT” cocktail. time for rational and safe alternatives. Pediatr Clin North Am 1989; 36:1285–91.
6. American Academy of Pediatrics Committee on Drugs. Reappraisal of lytic cocktail/Demerol, Phenergan, and Thorazine (DPT) for the sedation of children. Pediatrics 1995; 95:598–602.
7. White PF, Way WL, Trevor AJ. Ketamine: its pharmacology and therapeutic uses. Anesthesiology 1982; 56:119–36.
8. Levanen J, Makela M, Scheinin H. Dexmedetomidine premedication attenuates ketamine-induced cardiostimulatory effects and postanesthetic delirium. Anesthesiology 1995; 82:1117–25.
9. Gutstein HB. Potential physiologic mechanism for ketamine-induced emergence delirium [letter]. Anesthesiology 1996; 84:474.
10. Morray JP, Lynn AM, Stamm SJ, et al. Hemodynamic effects of ketamine in children with congenital heart disease. Anesth Analg 1984; 63:895–9.
11. Audenaert SM, Wagner Y, Montgomery CL, et al. Cardiorespiratory effects of premedication for children. Anesth Analg 1995; 80:506–10.
12. Roerig DL, Kotrly KJ, Vucins EJ, et al. First pass uptake of fentanyl, meperidine, and morphine in the human lung. Anesthesiology 1987; 67:466–72.
13. Lockhart CH, Nelson WL. The relationship of ketamine requirement to age in pediatric patients. Anesthesiology 1974; 40:507–8.
14. Blumer JL. Clinical pharmacology of midazolam in infants and children. Clin Pharmacokinet 1998; 35:37–47.
15. Feld LH, Negus JB, White PF. Oral midazolam preanesthetic medication in pediatric outpatients. Anesthesiology 1990; 73:831–4.
16. McMillan CO, Spahr-Schopfer IA, Sikich N, et al. Premedication of children with oral midazolam. Can J Anaesth 1992; 39:545–50.
17. Tobias JD, Phipps S, Smith B, Mulhern RK. Oral ketamine premedication to alleviate the distress of invasive procedures in pediatric oncology patients. Pediatrics 1992; 90:537–41.
18. Gutstein HB, Johnson KL, Heard MB, Gregory GA. Oral ketamine preanesthetic medication in children. Anesthesiology 1992; 76:28–33.
19. Bachenberg KL. Oral ketamine for the management of combative autistic adult [letter]. Anesthesiology 1998; 89:549.
20. Steinbrook RA, Hughes N, Fanciullo G, et al. Effects of alkalinization of lidocaine on the pain of skin infiltration and intravenous catheterization. J Clin Anesth 1993; 5:456–8.
21. Lebovic S, Reich DL, Steinberg G, et al. Comparison of propofol versus ketamine for anesthesia in pediatric patients undergoing cardiac catheterization. Anesth Analg 1992; 74:490–4.
22. Parish JM, Shepard JW. Cardiovascular effects of sleep disorders. Chest 1990; 97:1220–6.
23. Hamza J, Ecoffey C, Gross JB. Ventilatory response to CO2 following intravenous ketamine in children. Anesthesiology 1989; 70:422–5.
24. American Academy of Pediatrics Committee on Drugs. Guidelines for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures. Pediatrics 1992; 89:1110–5.
25. American Society of Anesthesiologists Task Force on Sedation and Analgesia by Non-Anesthesiologists. Practice guidelines for sedation and analgesia by non-anesthesiologists. Anesthesiology 1996; 84:459–71.
26. Kohrs R, Durieux ME. Ketamine: teaching an old drug new tricks. Anesth Analg 1998; 87:1186–93.
27. Gingrich BK. Difficulties encountered in a comparative study of orally administered midazolam and ketamine [letter]. Anesthesiology 1994; 80:1414–5.
© 2000 International Anesthesia Research Society